Congratulations to Min Zhang for passing her defense on Dec 9, 2022. Her PhD dissertation is titled with “Explainable and Green Solutions to Point Cloud Classification and Segmentation”. Her dissertation Committee members include Prof. C.-C. Jay Kuo (Chair), Keith Jenkins, and Prof. Stefanos Nikolaidis (Outside member). Min’s presentation was highly praised by the Committee. We invite Min Zhang here to share an abstract of her thesis and her defense experience. We wish Min Zhang all the best for her future career and life!
Point cloud processing is a fundamental but challenging research topic in the field of 3D computer vision, we specifically study two point cloud processing related problems — point cloud classification and point cloud segmentation. Given a point cloud as the input, the goal of classification is to label every point cloud as one of the object categories and the goal of segmentation is to label every point as one of the semantic categories. State-of-the-art point cloud classification and segmentation methods are based on deep neural networks. Although deep-learning-based methods provide good performance, their working principle is not transparent. Furthermore, they demand huge computational resources (e.g., long training time even with GPUs). Since it is challenging to deploy them in mobile or terminal devices, their applicability to real world problems is hindered. To address these shortcomings, we design explainable and green solutions to point cloud classification and segmentation.
We first propose an explainable machine learning method, PointHop, for point cloud classification and further improve its model complexity and performance in PointHop++. Then, we extend the PointHop method to do explainable and green point cloud segmentation. Specifically, an unsupervised feedforward feature (UFF) learning scheme for joint classification and part segmentation of 3D point clouds and an efficient solution to semantic segmentation of large-scale indoor scene point clouds (i.e., the GSIP method) are proposed. Finally, we investigate local and global aggregation in point cloud classification and segmentation and propose SR-PointHop for green point cloud classification using single resolution representation and extensive geometric aggregation, and GreenSeg for segmenting both small-scale and large-scale point clouds efficiently and effectively with a green local aggregation strategy.
First, I would like to thank Professor Kuo and all lab members for helping me during my PhD journey. Professor Kuo supports me a lot during the hard times caused by COVID19. I am always motivated and encouraged by his enthusiasm and diligent attitude to research. All the MCL lab members are very kind, they are never afraid to share their experience and knowledge with each other. I learned a lot from them and thanks for being my good role models. All of you are my life-long friends and I really cherish this. PhD study is a short period in our life, we will make it in the end. Never give up! I wish all the best luck to our family members for a great and bright future.
— Min Zhang
